Table 4.

Comparison of the average performance metrics from five-fold cross-validation for different classifiers and the stacking classifier.

	Overall	Weighted with 95% CI
Classifier	Accuracy	Precision	Recall	F1-score	Specificity
Linear Discriminant Analysis (LDA)	67.88 ± 2.27	67.69 ± 2.27	67.88 ± 2.27	67.88 ± 2.27	67.77 ± 2.27
XGBoost (XGB)	81.43 ± 1.89	81.37 ± 1.89	81.43 ± 1.89	81.43 ± 1.89	81.39 ± 1.89
Random Forest (RF)	82.91 ± 1.83	82.87 ± 1.83	82.91 ± 1.83	82.91 ± 1.83	82.74 ± 1.84
Logistic Regression (LR)	68.37 ± 2.26	68.63 ± 2.26	68.37 ± 2.26	68.37 ± 2.26	68.47 ± 2.26
Support Vector Machine (SVM)	62.28 ± 2.36	70.03 ± 2.23	62.28 ± 2.36	62.28 ± 2.36	61.53 ± 2.37
AdaBoost	74.66 ± 2.12	74.45 ± 2.12	74.66 ± 2.12	74.66 ± 2.12	74.22 ± 2.13
K-Nearest Neighbors (KNN)	79.17 ± 1.97	79.11 ± 1.98	79.17 ± 1.97	79.17 ± 1.97	79.13 ± 1.98
Gradient Boosting (GB)	89.88 ± 1.47	89.86 ± 1.47	89.88 ± 1.47	89.88 ± 1.47	89.87 ± 1.47
Stacking model (GB + RF + XGB)	91.45 ± 1.36	91.44 ± 1.36	91.45 ± 1.36	91.45 ± 1.36	91.44 ± 1.36